[Paper Review] High-throughput screening and mechanistic insights into solid acid proton conductors

Jonas Hänseroth, Max Großmann+3 more|arXiv (Cornell University)|Feb 16, 2026

Machine Learning in Materials Science1 citations

TL;DR

The paper introduces a two-stage, machine-learning–assisted high-throughput workflow to directly compute proton diffusion in solid acids, screening over six million materials and identifying 27 top candidates with mechanistic insight, including a universal ~2.5 Å oxygen–oxygen transfer distance across diverse chemistries.

ABSTRACT

Proton-conducting solid acids could enable water-free operation of high-temperature fuel cells. However, systematic materials screening has, hitherto, been computationally prohibitive. Here, we introduce a two-stage high-throughput screening strategy that directly computes proton diffusion coefficients, enabled by machine-learned interatomic potentials fine-tuned to ab initio data. Starting from more than six million materials, our screening -- based on structural motifs rather than empirical descriptors -- identifies $27$ high-performing proton conductors, including over ten previously unexplored compounds. These include sustainable and commercially available materials, candidates that have not yet been synthesized, organic systems that fall outside conventional design rules, and known proton conductors that validate our approach. Importantly, our findings reveal a universal oxygen--oxygen distance of approximately $2.5$~Å at the moment of proton transfer across diverse chemistries, providing mechanistic insight and showing that macroscopic proton conductivity emerges from the interplay between anion rotational dynamics, hydrogen-bond network connectivity, and proton-transfer probability.

Motivation & Objective

• Aim to enable water-free, high-temperature proton conduction by identifying solid acid materials with high proton mobility.
• Develop and deploy a motif-based structural screening strategy to filter candidates from huge materials databases.
• Combine fast ML-driven MD screening with ab initio refinement to obtain reliable proton diffusion coefficients.
• Elucidate the mechanistic relationship between anion dynamics, hydrogen-bond networks, and proton transfer in solid acids.

Proposed method

• Screen >6 million materials from Materials Project and Alexandria databases using two structural motifs inspired by CsH2PO4 and Cs7(H4PO4)(H2PO4)8 to identify oxoanion frameworks capable of proton transfer.
• Run 100 ps MD simulations with the MatterSim foundation model to filter dynamically stable candidates and rank proton transport potential.
• Perform 40 ps AIMD on the top 70 candidates to generate material-specific data for fine-tuning MACE MLIPs.
• Use fine-tuned MACE models to run 3 ns MD simulations at 400–600 K to obtain quantitative diffusion coefficients and anion rotation metrics.
• Apply path-integral MD (PIMD) for selected cases to assess nuclear quantum effects on transfer distances.
• Analyze oxygen–oxygen distance distributions and proton transfer event statistics to uncover universal transfer geometries.

Experimental results

Research questions

• RQ1 Can a motif-based, high-throughput screening workflow identify solid-acid proton conductors with high diffusion coefficients from a massive materials space?
• RQ2 What are the mechanistic factors (anion rotation, hydrogen-bond connectivity, transfer probability) that govern macroscopic proton conductivity in solid acids?
• RQ3 Is there a universal structural feature governing proton transfer distance across diverse oxoanionic chemistries?
• RQ4 How do nuclear quantum effects influence proton transfer distances in these materials?

Key findings

• Identified 27 top solid-acid proton conductors with high predicted diffusion coefficients, including both known conductors and unexplored compounds.
• Two-stage screening (motif filtering followed by MLIP-driven MD and AIMD refinement) efficiently narrows six million candidates to a focused set.
• Proton diffusion coefficients at 400 K for top candidates exceed 0.003 Å2 ps−1 in several cases, comparable to or better than CsH2PO4-based systems.
• Anion rotational dynamics and hydrogen-bond network connectivity modulate macroscopic diffusion; faster anion rotation does not always yield higher conductivity, indicating optimal balance is needed.
• A universal oxygen–oxygen transfer distance of ~2.5 Å governs proton hopping across diverse chemistries, confirmed by AIMD and recovered in PIMD for Te-containing outliers.
• The workflow recovered several experimentally validated conductors (e.g., CsHSO4, KHSO4, KH2PO4) and suggested new candidates including organic salts and earth-abundant compositions.